Page 1 of 1

LPM, LOGIT and PROBIT have all the same probability

Posted: Wed Nov 12, 2014 5:16 pm
by Econometrics Student
Hello guys,

I have a simple question to you,

I need to estimate three coeffficient (y=b1+b2x2+b3x3) that LPM, LOGIT and PROBIT have all the same probability. Do you know how I do that?

Thanks for your help,

Regards.

Re: LPM, LOGIT and PROBIT have all the same probability

Posted: Sun Nov 16, 2014 11:32 am
by Econometrics Student
Someone, please?

Re: LPM, LOGIT and PROBIT have all the same probability

Posted: Sun Nov 16, 2014 11:52 am
by startz
I don't believe it's possible. Maybe you could explain further.

Re: LPM, LOGIT and PROBIT have all the same probability

Posted: Mon Nov 17, 2014 4:01 am
by Econometrics Student
I estimated for a simple model (y=b1+b2x2+b3x3) LPM, LOGIT and PROBIT.

After that, i have to estimate a model that: P is the probability of the event "Y=1 "estimated under the condition:
be identical for all individuals in the sample. So, I have to determine the estimates of the coefficients in each of the three classes of such models.

Can you help me?

Thank you.

Re: LPM, LOGIT and PROBIT have all the same probability

Posted: Mon Nov 17, 2014 5:21 am
by trubador
I am not sure, but it sounds like you are asked to estimate a constant probability model that includes only the intercept term (i.e. y = b1).